QBrainNet: harnessing enhanced quantum intelligence for advanced brain stroke prediction from medical imaging

Area of Science:

• Quantum computing
• Medical imaging analysis
• Artificial intelligence in healthcare

Background:

• Brain stroke is a leading cause of death and disability worldwide, necessitating early and accurate diagnosis.
• Classical machine learning models like Convolutional Neural Networks (CNNs) face performance limitations with complex, high-dimensional medical image data.
• Existing methods struggle to capture intricate, non-linear patterns crucial for effective stroke detection.

Purpose of the Study:

• To introduce QBrainNet, a novel quantum-enhanced model for improving brain stroke prediction from medical imaging.
• To leverage quantum properties for superior feature extraction and classification accuracy in stroke detection.
• To demonstrate the potential of quantum computing in revolutionizing medical diagnostics.

Main Methods:

• A hybrid classical-quantum approach involving classical preprocessing and quantum-enhanced learning.
• Utilizing Quantum Neural Networks (QNNs) with quantum properties like superposition and entanglement for feature extraction.
• Employing Variational Quantum Circuits (VQCs) to optimize classification by tuning quantum gates and operators.

Main Results:

• QBrainNet achieved a superior accuracy of 96% and an AUC-PR of 0.97.
• Demonstrated significantly enhanced prediction accuracy and efficiency compared to classical models (CNN, SVM, Random Forest).
• The quantum-enhanced model effectively captures complex, non-linear patterns in medical images related to stroke.

Conclusions:

• QBrainNet shows superior performance in stroke detection, outperforming established classical models.
• The model's shorter inference time makes it suitable for real-time clinical applications.
• Quantum computing holds transformative potential for medical diagnostics, particularly in stroke prediction.